Week 3 Lectures 5 & 6
Popular in Contemp Econ Issues >2
Popular in Department
This 3 page Class Notes was uploaded by Taryn manciu on Tuesday October 18, 2016. The Class Notes belongs to Econ 101 at University of Oregon taught by Urbancic M in Fall 2016. Since its upload, it has received 13 views.
Reviews for Week 3 Lectures 5 & 6
Report this Material
What is Karma?
Karma is the currency of StudySoup.
Date Created: 10/18/16
EC101 Week 3 Lecture 1 Samples with built in bias - Asked Oregon students which school is the best - HTLWS example: polls said Alf would win, but the problem with that was the survey was taken by phone and not everyone had a telephone Online Polls - Specific audience - Externalities - People sending people to polls makes them meaningless - Margin of error is present in online voting The Sample Matter - People self-selecting into online polls is exactly why they’re meaningless - The sample is not going to be representative - Its like the Oregon duck asking who has the best mascot and voting as many times as you want… Different Perspectives - Old vs. young: wage increases, college tuition, women’s rights, and social security (these are all issues that older and younger generations would have differing opinions about) Data: Plural Datum: Singular Data can be messy… - Guessing how many pennies are in a jar - Answers given: 246.1 does this mean $24.61, $246.10 was the decimal in the wrong place so its actually $2461? Or is it the amount of pennies in the jar 2,461? o The fact of the matter is, is that someone has to make these kind od decisions when going through all the data being collected o Someone has to accept or deny these figures o Things that also need to be taken into account § What was the response rate? § Are values missing? § Are some data being dropped? § Does this person dropping data have a specific agenda? Distributions - Normal Distributions o Bell shaped o Mean median and mode are quite close OR exactly the same o Normal Dist. Ex: height of males - Negatively Skewed o Long tail towards left o Ex; grades, age of death - Positively Skewed o Long tail skewed towards the right o Ex: income, wealth, house price Week 3 Lecture 2 Little Figures That Are Not There - FIRST problem with studies with small sample sized the results may vary quite a bit between them (or between each small study and one with a larger sample) - This is why small studies aren’t reputable - SECOND problem with studies with small sample sized is that many of these studies might be done, and it may be the case that only the one with the most favorable results is being reported (while the rest are sent to the trash) o This makes things very deceptive Much Ado About Nothing - The absolute number look big but the context they are in make them tiny o $562,000 is 0.0000014% of the budget of the department of veterans affairs… TINY - Be wary whenever an absolute number is given without the larger context o Is this number big for its context? o Is it small for its context? o Does it matter? - Ponder weather a percentage is meaningful o Is the % change significant? o Does it fall with a normal range or variation? o Does the absolute number matter? LISTEN UP: MEASUREMENTS DO NOT MEAN TRUTH - When data are collected to measure something in the real world there is always some measurement error. ALWAYS for any discipline - ESSPECIALLY true in social sciences o Political campaigns - If the gap between two measured values is less than the margin of error we cannot really sat t that they are different or that one is larger than the other. The Gee-Whiz Graph - Beware of graphs that exaggerate effects by manipulating the scale - Having inconsistent scale - Omitting the scale altogether